Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
5 CONCLUSIONS AND RECOMMENDATIONS CONCLUSIONS 1. Compared with 4 years ago, the Army has consider- ably improved its awareness and its implementation of AI and robotics development programs. These projects cover a restricted range of applications, including robot mobile vehicles, both teleoperated and autonomous; material handling robot systems; and a number of expert systems. Although many more applications that have been identified are not funded, the funded projects could be very useful for initiating and implementing future systems. 2. The Army has wisely concentrated its financial and managerial resources on a small number of projects. Although this is partly in agreement with our 1983 recommendation, the time period will be longer than the 2 to 3 years we had hoped for. Therefore, more funding will be needed to bring these projects to use in the field. We still believe that it is important to get early introduction of successful, simple robotic systems. 3. Robotics and AI still seem not to have visibility or champions at the highest levels in the Army. These technologies have tremendous potential to reduce hazards, increase individual effectiveness, reduce costs, and provide needed high-technology improvements in military and logistic operations. The establishment of AI and robotics technology base groups and the compilation of a master plan by TRADOC may reflect progress toward solving this problem. Without high-level support, the projects will not receive sufficient funding. 24
25 4. Education at all levels is the Army's greatest need as it adopts robotics and AI. With the growing, more integral role of the Armed Services in the nation's research and development effort, the Army needs to enhance its ability to attract and retain key personnel, both civilian and military. Career-long education and training is the key to a specialized Army work force that provides scientific and technological leadership, during peace and war. 5. The Army's work in robotics and artificial intelli- gence needs periodic external review and oversight. The Army Science Board is an obvious choice to be the focal point, but the task appears to require too much work for the board itself. The board, however, could pull together the people needed to perform the reviews. 6. Technical issues that the Army needs to address are: Communication links for teleoperation, Stereoscopic vision, Map reading, Sensors for target acquisition, Deflection-compensated robot arms, Stabilizing systems, and Small, self-contained power sources. 7. The Army has paid insufficient attention to utiliz- ing existing robot technology for performing repetitive tasks in rear-echelon posts, depots, and warehouses. These tasks could include loading, unloading, and packag- ing. Mounting existing programmable robots on vehicles such as trucks, jeeps, or wheeled platforms would provide means of performing repetitive tasks with minimum training required if existing hand-held interface boxes that offer "training by doing" were used. Many tasks do not need customized robots that conform to very expensive specifi- cations for battlefield survivability. Use of available technology would significantly decrease the time required to familiarize large numbers of soldiers with robots and their capabilities and programming and for the initiation of many new applications by the creative few. 8. Long-term planning and commitments are missing from Army AI and robotic educational programs. Evidence of this is that continued support for the AI Centers at the
26 University of Pennsylvania and the University of Texas is not certain, and a proposal for an Army AI center was dropped. RECOMMENDATIONS 1. Increase the financial support for near-term robotic applications: Teleoperated Mobile Anti-Armor Project, Field Material Handling Robot, Soldier/Robotic Interface Prject. It appears possible to have field- useful systems within 5 years if sufficient funds are made available. In particular, concentrate attention on the TMAP project, with greatly increased emphasis on the development of counter-countermeasures and secure, survivable, non-jammable data links from the outset. 2. Establish a central, high-level Army organizational unit responsible for integrating and monitoring all AI and robotics R&D programs and guiding the subsequent procurement of fieldable systems. 3. Mount a serious effort to develop secure, sur- vivable counter-countermeasures or secure, reliable, non-jammable, multidirectional communications (data links). These complex problems must be addressed from the outset--and solved--if the systems sought are to possess true operational utility. 4. Initiate studies aimed at the exploitation of existing legged mobile robot technology. The potential combination of deflection-compensating robot arms and automatic attitude stabilization in rough terrain could significantly advance TMAP application. 5. Initiate far more comprehensive educational and training programs for officers and technicians in AI and robotics, emphasizing the former. 6. Establish one or more university centers of excellence for AI and robotics, as part of the programs sponsored by the Army Research Office. Because these technologies are of interest to all the military services, co-sponsorship of such centers, or coordination under the University Research Initiative or other program, would be desirable.
27 7. Initiate a short study to determine the potential utility of equipping an existing Army truck or other vehicle with a commercially available, programmable robot for loading/unloading operations at depots or in the field.
